diff --git a/exercises/practice/anagram/.docs/instructions.append.md b/exercises/practice/anagram/.docs/instructions.append.md
index 8d71a920..273ca3cb 100644
--- a/exercises/practice/anagram/.docs/instructions.append.md
+++ b/exercises/practice/anagram/.docs/instructions.append.md
@@ -1,3 +1,5 @@
 # Instructions Append
 
+## Implementation
+
 The anagrams can be returned in any order.
diff --git a/exercises/practice/complex-numbers/.docs/instructions.append.md b/exercises/practice/complex-numbers/.docs/instructions.append.md
index 1f221d62..cb6699fe 100644
--- a/exercises/practice/complex-numbers/.docs/instructions.append.md
+++ b/exercises/practice/complex-numbers/.docs/instructions.append.md
@@ -1,4 +1,6 @@
 # Instructions append
 
+## Implementation
+
 You also need to implement a `equal/2` function. For this you can consider
 two numbers as equal, when the difference of each component is less than 0.005.
diff --git a/exercises/practice/hamming/.docs/instructions.append.md b/exercises/practice/hamming/.docs/instructions.append.md
index 58c49ec1..e50b4e99 100644
--- a/exercises/practice/hamming/.docs/instructions.append.md
+++ b/exercises/practice/hamming/.docs/instructions.append.md
@@ -1 +1,5 @@
+# Instructions append
+
+## Implementation
+
 If the sequences are not of equal length, return a tuple of `{error, badarg}`. Otherwise, return only the integer.
diff --git a/exercises/practice/robot-simulator/.docs/instructions.append.md b/exercises/practice/robot-simulator/.docs/instructions.append.md
index bcda7f78..c7d8e12b 100644
--- a/exercises/practice/robot-simulator/.docs/instructions.append.md
+++ b/exercises/practice/robot-simulator/.docs/instructions.append.md
@@ -1,5 +1,7 @@
 # Instructions append
 
+## Implementation
+
 This exercise does require some kind of global state for your
 robot. In erlang this can be done in various ways (starting with most
 idiomatic one):